|
|
|
|
MOR Journal Abstracts
|
 | |
| |
| |
|
Dynamic
Routing of Unmanned Aerial Vehicles Using Reactive Tabu Search (Kevin P.
O'Rourke, William B. Carlton, T. Glenn Bailey and Raymond R. Hill)
Unmanned Aerial Vehicle (UAV) routing is a complex
problem faced by the operators of long-duration, unmanned aerial vehicles such
as the US Air Force’s RQ-1A Predator. The
US Air Force uses the Predator UAV to perform reconnaissance and surveillance
missions. The Predator is
remotely flown by Air Vehicle Operators, who are Air Force pilots, located in
a Ground Control Station. Existing
mission support routing software automatically generates deterministic items such as terrain avoidance profiles, ground
station to UAV line of site availability, route times between defined way
points, fuel consumption, heading and turn information, etc., but it does not
and will not optimize route selection over assigned targets.
This routing task is left to the operator.
Even basic routing problems are computationally difficult to solve.
Unfortunately, real-world problems such as UAV routing possess many additional
side constraints that impose route and vehicle capacity limitations, route
length bounds, and feasible visitation time window restrictions.
Additionally, the routing task may employ a variety of vehicles located
at various operating locations. Fortunately, the tabu search heuristic provides excellent results
on these types of problems. In
this article, we present the application of a reactive tabu search
metaheuristic to the vehicle routing problem with time windows using the
object-oriented Java programming language. We describe the specifics of the
reactive tabu search, additional considerations for the UAV routing
application, and present computational results of our RTS using a benchmark
set of routing problems and a notional UAV operational application.
This paper was awarded the MORS 2000 Barchi Prize. (Pg. 5)
Exploring
Threat Intent Measures for System Evaluation (Michael D. Proctor, Lou
Lartigue and Ephraim Martin)
Exploring Threat Intent Measures for System Evaluation evolved from an effort to evaluate manned reconnaissance systems in a combat simulation. Specifically, the TRACER program is a joint US/UK development program to field the next generation scout vehicle. The vehicle is projected to cost on the order of 3 to 7 million dollars per copy. Lockheed Martin Missiles and Fire Control in conjunction with several partners developed a technical proposal for a concept development effort for TRACER. The value of target information as viewed in context by a human is typically not captured in a combat simulation. This value is generally related to the persistence and quality of the human observation. The article explores measures of this military intelligence meaningful to field commanders involved in the acquisition. (Pg. 31)
Assessing
the Financial Condition of Civil Reserve Air Fleet Participants
(William (Bud) F. Bowlin)
The
Civil Reserve Air Fleet (CRAF) is a key component of the Department of
Defense’s (DoD) air mobility capability.
In the early 1990s, several CRAF participants liquidated their
operations due to bankruptcy and other participants entered bankruptcy but did
not liquidate. Consequently, the
DoD lost some CRAF participants
and was in danger of losing others. William
(Bud) F. Bowlin uses a non-traditional financial analysis tool, data
envelopment analysis, to assess the financial condition of CRAF participants.
(Pg. 43)
Retention
and Reacquisition of Military Skills (Mark A. Sabol and Robert A. Wisher)
The
Army trains thousands of tasks, often months before they are needed.
But a century of memory research tells us that soldiers will lose much
between acquisition and utilization. This
report, voted best paper presented to the working group on Analytic Support to
Training at the 67th MORSS (1999), reviews what is known about the
forgetting of military tasks. It
presents a meta-analytic review of practical research by many scientists at
military and academic laboratories, followed by recent studies by the authors
on reacquiring skills once they are lost.
This work was done for the Army’s Training and Doctrine Command. (Pg.
59)
Lanchester’s
Equations And The Structure of The Operational Campaign: Within-Campaign
Effects (Ron Speight)
This
article considers the application of Lanchester’s attrition equations to
operational campaigns and battles. Lanchester theory has had an influence,
both on the way that attrition is represented in aggregated battle models, and
on beliefs about the likely effects of the concentration of force. However,
the Lanchester formulation effectively ignores the structure that may be
imposed on the battle by the command process. This article considers the
manner in which this structure may affect the workings of Lanchester theory.
After
briefly outlining the main features of Lanchester theory and outlining the
scope of this investigation, the paper examines possible motives for fitting
Lanchester-derived formulae to historical battle data. It suggests that the
main motive is to establish the strength and form of any statistical
relationship between the sizes of the opposing forces on the one hand, and the
likely balance of attrition on the other. It then discusses some structural
factors that may affect the workings of Lanchester theory. After some
numerical and qualitative analysis of the Iwo Jima, Inchon-Seoul and Ardennes
campaigns, it reviews the results in the light of the factors just alluded to.
Based on this evidence it suggests that Lanchester theory is unlikely to
provide a good account of the balance of attrition and its likely time profile
in any but the simplest of operational campaigns. If accurate description is
the aim then the analyst will have to consider: the overall objectives of the
commanders; the likely orchestration demands of the campaign; the
opportunities for the concentration of force (and the non-terrain factors
which may facilitate such concentration); the dynamic assignment of missions
and the likely adaptive reactions; loss mechanisms other than those covered by
conventional attrition; and, above all, the likely effectiveness of the forces
involved. (Pg. 81)
An
IO Conceptual Model and Application Framework (Patrick Allen and Chris
Demchak)
In
1999, DARPA sponsored an SBIR study to develop the conceptual basis for
simulations of Information Warfare (IW) to be used in support of training
events. The results of this study
produced a conceptual model to provide the framework for future IW and
Information Operations (IO) models and simulations, and described the need
for, and design of, an IO Federation of Models.
It also discussed how to provide multilevel security during a training
exercise. This conceptual model
has been used both by GDES and other contractors to guide the development of
future IO/IW models, and has gained international interest in future work.
(Pg. 9)
Risks
of Cyber Attack to Water Utility Supervisory Control and Data Acquisition
Systems (Barry C. Ezell, Yacov Y. Haimes and James H. Lambert)
Supervisory Control and
Data Acquisition (SCADA) systems control US water
supply and treatment processes. The
President’s Commission on Critical Infrastructure identified cyber terrorism
as a threat to our critical infrastructure.
The commission stated that waiting for a [cyber] disaster is a
dangerous strategy. Barry Ezell,
Yacov Haimes and Jim Lambert developed a probabilistic risk assessment and decision-making methodology to assess
and manage the risks of willful threats to water utility SCADA systems.
The framework may assist decision-makers in understanding the risks of
cyber intrusion, and their consequences and associated tradeoffs, in order to
improve the survivability of the system. (Pg. 19)
The
Network Diversion Problem (Norm
Curet)
The
network diversion problem can be viewed as an information operations network
flow application, whereby the most advantageous combinations of nodes and
links are removed from a given network so as to divert the information flow
over a specified set of links.
The problem can be formulated as
a mixed integer programming model, but for large network sizes, the solution
proves to be a formidable task. A
lagrangean relaxation decomposition algorithm is proposed that exploits the
underlying network flow structure, enabling problems containing thousands of
nodes and links to be efficiently solved. (Pg. 31)
Valuing
Psychological Operations (Philip M. Kerchner, Richard F. Deckro and Jack M.
Kloeber Jr.)
“Soft
Targeting” in information operations is becoming an increasingly important
consideration. Psychological
Operations (PSYOP) are one of the classic approaches to targeting an
opposition’s will to fight. While PSYOPs are a long-standing military
practice, there have been no fixed measures of merit for evaluating PSYOP
options. This study reports on using Value-Focused Thinking and
multi-objective value analysis to develop first cut measures for the
evaluation of PSYOP products. The
value hierarchy, based on doctrine and expert opinion, is reviewed. Its use is illustrated via the analysis of a set of PSYOP
options in a notional scenario. Several
changes in the way PSYOP products are evaluated and considered are identified.
The model allows the Psychological Operations Detachment Commander to quantify
the potential of proposed PSYOP products in meeting the theatre commander's psychological
objectives. (Pg. 41)
Military
Operations Research Society Oral History Project Interview of Eugene
P. Visco, FS (Eugene P. Visco, FS, Robert Sheldon and Jack Marriott
(Pg.
63)
The
Deployment Analysis Network Tool Extended (DANTE) (Michael K. Williams,
Owen Spivey and Thom J. Hodgson)
The Deployment Analysis Network Tool Extended (DANTE) is an analysis tool for studying large-scale deployment scenarios. DANTE was developed originally to support analysis of strategic mobility concepts in an Army Science Board study, and is used by the Military Traffic Management Command - Transportation Engineering Agency to help evaluate alternatives in the Army redesign. DANTE represents a deployment from bases in the United States, from forward-based locations and from pre-positioned stocks as a time-phased network flow. The objective is to minimize the time to close the deploying force on the staging area. A graphic user interface facilitates the modification of deployment parameters so the impacts of changes on the time to close can be quickly determined. (Pg. 5)
Multiple
Perspective R&D Portfolio Analysis for the National Reconnaissance
Office's Technology Enterprise (Gregory S. Parnell, Benjamin I. Gimeno,
Deborah Westphal, Joseph A. Engelbrecht and Richard Szafranski)
The
National Reconnaissance Office (NRO) Technology Enterprise desired to make
R&D portfolio decisions by assessing how their portfolio supports the
needs of future customers. We
surveyed customers to understand their future needs for intelligence
information. Based on customer needs, we qualitatively and quantitatively
defined the potential future value of intelligence capabilities.
The NRO Technology Enterprise leadership specifically requested the
development of a value model with audacious
objectives that would challenge their managers and technologists. The
portfolio was analyzed from multiple perspectives: future customer value,
R&D strategy, functional area balance, value gaps and time. The multiple
perspective analysis gave insights not normally obtained in typical portfolio
analyses that examine benefit and cost. (Pg. 19)
Biological Warfare Human Response
Modeling (Gillian L. Rickmeier, Gene E. McClellan and George A. Anno)
Casualty estimation is an important tool
for military and medical planning. The Knowledge Acquisition Matrix Instrument
(KAMI) analyzes the effects of bioagent-induced diseases that are wartime or
terrorist threats but for which only limited human response data is available.
The article uses the KAMI technique to supplement available data with input
from subject matter experts to predict time-phased casualty estimates
resulting from anthrax, botulism, pneumonic plague and Venezuelan equine
encephalitis. Ms. Rickmeier, Dr. McClellan, and Mr. Anno created KAMI to
develop casualty tables for Allied Medical Publication 8, a medical planning
guide. This work was awarded the best presentation of Working Group 23 at the
annual MORS Symposium in 1999. (Pg. 35)
Using
Simulation to Model Time Utilization of Army Recruiters (James D. Cordeiro,
Mark A. Friend, J.O. Miller, Kenneth W. Bauer, Jr., Jack M. Kloeber)
Over
the last few years, all of the US military services (with the exception of the
Marines) have been struggling to meet recruiting quotas.
The Army in particular has seen a significant increase in their yearly
quotas without a corresponding increase in personnel authorizations at the
recruiting stations. This paper starts with basic problem definition and details
the steps involved in conducting a simulation study designed to provide
Headquarters US Army Recruiting Command with a model to examine the effects of
various policy decisions on recruiting station output. Selected as best
presentation in the Manpower and Personnel Working Group at MORS 1998 and
nominated for the Barchi Prize. (Pg. 59)
A
Bayesian Stochastic Formulation of Lanchester Combat Theory (Abdul Sattar
Rashid Salim and W. M. Hamid)
In recent years the advancements in defense planning have created many challenging problems (deterministic and stochastic) for the scientist in the operations research area. Some of these challenges deals with the mechanism of incorporating field commanders subjective information and engineers prior assessments (weapons reliability) in the decision process. Furthermore, data acquisition in the probabilistic models should be admissible and non-prohibitive. This article addresses these two objectives, and presents a stochastic formulation that incorporates experts' opinion and requires feasible (easy and fast) information from the battlefield to assist the field commanders to review their strategies as often as they desire. (Pg. 69)
Signals
from Space: The Next-Generation Global Positioning System
(Lee Lehmkuhl, David
Lucia and James Feldman)
The
GPS Joint Program Office (JPO) needed to recommend a new signal structure for
the next generation of GPS satellites to the Independent Review Team.
Lee Lehmkuhl, David Lucia and Ken Feldman used decision analysis to
model the value of GPS to different user communities and quantify tradeoffs.
The results allowed the GPS Independent Review Team to choose a new
signal with superior military value that also meets all civilian technical
performance requirements. (Pg. 5)
The Diesel Submarine Flaming Datum Problem (Alan Washburn and Ryusuke Hohzaki)
In the original World War II Flaming Datum Problem, a diesel
submarine has just torpedoed a ship and thereby reveled its own position.
The submarine’s problem is now to leave the area before pursuing
forces are able to find and attack her. The
situation can be thought of as a two-person zero-sum game.
The problem has been extensively studied since WWII, usually with the
constraints that the submarine cannot exceed a certain speed and that the
pursuing forces have only a limited endurance.
A diesel submarine is also limited by the energy stored in its
batteries, and this article analyzes the situation recognizing that fact.
The new constraint considerably complicates the analysis, so we are
able to achieve only bounds on the game value.
The bounds are close, however, and the bounding strategies are
interesting. The submarine’s
speed should be a continuously decreasing function of time that results in a
farthest-on distance of y(t) at time t.
The pursuing forces search uniformly over a circle with radius y*(t),
with y(t)<y*(t). Either bounding strategy could be easily implemented.
Among other results, we find that the endurance of the pursuing forces
is more important under the new rules. (Pg. 19)
Suppression And The Theory of Stochastic Duels
(Michael J Armstrong)
We consider a variation of a stochastic duel; that is, a firefight between two opposing tanks
which meet on the battlefield and begin shooting at each other.
Each time a tank fires, there are two obvious results that may occur:
the shot might destroy the opposing tank and so end the battle; or the shot
may miss, in which case the battle continues.
There is however a third possible result: a shot may fail to destroy
the target, but still temporarily render it incapable of returning fire by suppressing
it. Suppression is an important
and real (though transient) combat phenomena and has been the subject of
numerous empirical studies but has rarely been incorporated in theoretical
models. The purpose of this paper
is to analyze a stochastic duel model that incorporates this suppressive
effect of firepower.
Understanding Instability in a Complex Deterministic Combat
Simulation (Kevin J. Saeger and James H. Hinch)
In late 1996, analyses for the Quadrennial Defense Review were in full swing. James Hinch and Kevin Saeger were asked to develop a quantitative link between force effectiveness, force structure and system modernization. The authors employed a “tried and true” deterministic simulation to obtain measures of effectiveness and were instantly confronted with a dilemma—the simulation was extremely sensitive to small changes to the input data. In one case, the authors found that randomly varying a unit’s maximum speed by 1 kilometer per hour caused the total losses in a scenario to vary by more than 10 percent. Over the last three years, Saeger and Hinch have developed a systematic method for examining this simulation's variability and have developed a procedure for extracting robust answers from this complex deterministic combat simulation. This article provides an analytical basis and supporting empirical evidence for a solution technique that can be extended easily to a wide class of simulations. (Pg. 43)
A
History And Analysis of Operations Other Than War: Ask Not “What Forces
Should We Deploy” But Rather “What Is The Problem?”
(Eugene P. Visco, FS)
Some
of our links are in portable document format (pdf); they will be indicated by
this symbol
|
